1. Technical Field
The present invention relates to a system and method for improved name matching using regularized name forms. More particularly, the present invention relates to a system and method for regularizing candidate names and query names based upon their particular culture origin, and identifying names whose corresponding regularized candidate names meet a matching threshold when compared against a regularized query name.
2. Description of the Related Art
A major difficulty in successfully matching personal names stored in a database with a user-provided name query arises when variant forms of the name are possible either through 1) spelling variation inherent to the language itself, or 2) through spelling variation that arises when the names are transliterated into the Roman alphabet from other writing systems.
One approach relies on phonetically based rewrite rules that convert a name to a phonetic form approximating its pronunciation, along with the calculation of a phonetic distance value between two name forms that are being compared. A challenge found, however, is that this approach is only valid in cases in which alternate spelling variations for names that sound similar are inherent to the language itself. Name variants that arise from different transliteration conventions may not show evidence of such similarity in pronunciation. Furthermore, generating phonetic variants and calculating their similarity is computationally very expensive, making it necessary to create a static, pre-processed database that may not be changed or updated in real time. When a new record is added or a rule is changed, the entire database must be regenerated, which renders such a system impractical for most users.
What is needed, therefore, is a system and method that effectively and efficiently improve name-matching capabilities for names with spelling variations and transliteration variations.